THE PURE LINE HYPOTHESIS. 553 



It may be therefore affirmed that there was a very distinct 

 tendency for small leaves to occur with dwarf habit and varieg^a- 

 tion. 



It is interestinLC to notice that we have here three degrees 

 of seg^regation : in varieg-ation the seo-reg^atioii was practically 

 perfect; in dwarf habit it was less perfect, and in size of leaf it 

 was quite imperfect, althoug-h distinguishable. 



On the chromos(.)mc hypothesis, and with relatively invari- 

 able factors, a practically perfect segregation is intelhgible ; but 

 it is not obvious as to how it is possible to interpret the occur- 

 rence of a very imperfect or barely recognisable segregation. If, 

 however, the notion of stable factors be abandoned, an imperfect 

 segregation could perhaps be explained by supposing that the 

 intensity of the mutually influencing action between the two oppos- 

 ing factors in the homologous chromosomes varied in the germ- 

 cells from which the different individuals arose, so that in some 

 a mean of the two original factors was attained, and in others 

 the two factors remained in a less altered condition. 



4. The Deep Yellow Chroiuoplasts. 



As already remarked, it cannot be stated that the deep 

 yellow chromoplasts were entirely absent in any of the flowers ; 

 i3Ut nearly always they were either abundant, being crowded 

 throughout the whole of the petals, or they were very scarce, and 

 mostly confined to the cells near the base of the petals. A few 

 more or less intermediate conditions were noticed where the 

 •chromoplasts were thinly scattered through the petals, but their 

 occurrence was quite exce]3tional among the plants under observa- 

 tion. Nevertheless, in a large bed of mixed nasturtiums, 

 recently examined in a public park, intermediate conditions 

 ajipeared to be less uncommon. 



We have seen that the male, dwarf progenitor had no 

 chromoplasts ; on self-fertilising the flowers, however, four off- 

 spring had chromoplasts and two had none. Similarly, the 

 female, straggling progenitor had no chromoplasts, but two of 

 the offspring possessed them and 18 did not. These erratic 

 numbers could possibly be ex]:)lained in Mendelian terms by pre- 

 supposing the existence of inhibitors or multiple factors, but the 

 material to hand is insufficient for arriving at very clear results. 



The 21 hybrids obtained by crossing these two plants com- 

 prised nine plants with chromoplasts and 12 without. The 

 families raised by self-fertilising the hybrids exhibited different 

 conditions. The actual results are given in the following table : 



